η6-Triphosphabenzene, η 5-triphosphacyclohexadienyl, and η 5-diphosphacyclopentadienyl complexes of group 8 and 9 metals: Heterocycle transformations at the metal center

Article abstract of DOI:10.1021/om0503489

The reactivity of a 1,3,5-triphosphabenzene, P₃C ₃Bu^t₃, and a triphosphacyclohexadienyl anion, [MeP₃C₃Bu^t₃]^-, toward a range of group 8 and 9 halide complexes has been investigated. The anion reacts with FeCl₂ to give the known tetraphosphaferrocene [Fe(η ⁵-l,3-P₂C₃Bu^t₃) ₂] and an unusual heterocage complex, [η⁵-1,3P ₂C₃Bu^t₃)Fe{P₂(PMe) ₂(CBu^t)₃}], via phosphinidene elimination and intramolecular phosphinidene transfer reactions, re-spectively. The reactions of P₃C₃Bu^t₃ or [MeP ₃C₃Bu^t₃]^- with [Cp*Ru(NCMe)₃] [PF₆] have afforded [Cp*Ru(η⁶-P₃C₃Bu^t ₃)] [PFe] and the first example of a triphosphacyclohexadienyl-transition metal complex, [Cp*Ru{η ⁵-(MeP)P₂C₃Bu^t₃}]. The latter can also be prepared by treating [Cp*Ru(η⁶-P ₃C₃Bu^t₃)] [PF₆] with MeLi. Similarly, the reaction of [η*Ru(η⁶-P ₃C₃-Bu^t₃)][PF₆] with LiA1H₄ has yielded the 2-H-triphosphacyclohexadienyl complex [Cp*Ru-{η⁵-P₃C₂Bu^t ₂(CHBu^t)}]. Treatment of [Rh(COD)Cl]₂ with PC₃Bu^t₃ in the presence of Na[BAr ^f₄], Ar^f = C₆H₃(CF ₃)₂-3,5, gave the complex [(COD)Rh(η ⁶-P₃C₃Bu^t₃)] [BAr ^f₄], which when reacted with either water or ethanol afforded the 1,1-addition products [(COD)Rh{η⁵-[(H)-(RO)P]P ₂C₃Bu^t₃}][BAr^f ₄], R = H or Et. The mechanisms of the various heterocycle transformation reactions are discussed, and the X-ray crystal structures of all new complexes are reported.

Full text of DOI:10.1021/om0503489

4216
Organometallics 2005, 24, 4216-4225
η6-Triphosphabenzene, η5-Triphosphacyclohexadienyl,
and η5-Diphosphacyclopentadienyl Complexes of Group
8 and 9 Metals: Heterocycle Transformations at the
Metal Center
Matthew D. Francis, Christian Holtel, Cameron Jones,* and Richard P. Rose
Center for Fundamental and Applied Main Group Chemistry, School of Chemistry,
Main Building, Cardiff University, Cardiff, CF10 3AT, U.K.
Received May 4, 2005
The reactivity of a 1,3,5-triphosphabenzene, P3C3But3, and a triphosphacyclohexadienyl
anion, [MeP3C3But3]-, toward a range of group 8 and 9 halide complexes has been
investigated. The anion reacts with FeCl2 to give the known tetraphosphaferrocene [Fe(η5-
1,3-P2C3But3)2] and an unusual heterocage complex, [(η5-1,3-P2C3But3)Fe{P2(PMe)2(CBut)3}],
via phosphinidene elimination and intramolecular phosphinidene transfer reactions, re-
spectively. The reactions of P3C3But3 or [MeP3C3But3]- with [Cp*Ru(NCMe)3][PF6] have
afforded [Cp*Ru(η6-P3C3But3)][PF6] and the first example of a triphosphacyclohexadienyl-
transition metal complex, [Cp*Ru{η5-(MeP)P2C3But3}]. The latter can also be prepared by
treating [Cp*Ru(η6-P3C3But3)][PF6] with MeLi. Similarly, the reaction of [Cp*Ru(η6-P3C3-
But3)][PF6] with LiAlH4 has yielded the 2-H-triphosphacyclohexadienyl complex [Cp*Ru-
{η5-P3C2But2(CHBut)}]. Treatment of [Rh(COD)Cl]2 with P3C3But3 in the presence of
Na[BArf4], Arf ) C6H3(CF3)2-3,5, gave the complex [(COD)Rh(η6-P3C3But3)][BArf4], which when
reacted with either water or ethanol afforded the 1,1-addition products [(COD)Rh{η5-[(H)-
(RO)P]P2C3But3}][BArf4], R ) H or Et. The mechanisms of the various heterocycle
transformation reactions are discussed, and the X-ray crystal structures of all new complexes
are reported.
Introduction
In addition, complexes containing η6-phosphinines are
well known, and examples have been utilized as cata-
lysts for the synthesis of pyridines from alkynes and
nitriles3 and the cyclodimerization of 1,3-butadiene.4
Moreover, treatment of phosphinines with alkyllithium
reagents yields 1-R-phosphacyclohexadienyl anions,
which in a few cases have been shown to participate in
salt elimination reactions with metal halides to give η5-
phosphacyclohexadienyl-transition metal complexes.1
Le Floch et al. have recently utilized this and other
routes to prepare phosphacyclohexadienylrhodium(I)
complexes, e.g., [Rh(COD)(η5-ButPC4Ph4)], and have
demonstrated their effectiveness as catalysts for the
selective hydroformylation of a range of olefins under
mild conditions.5
The coordination chemistry of ligand systems con-
taining low-coordinate λ3-phosphorus centers has rap-
idly expanded over the last two decades.1 It has become
clear that these ligands can display many similarities
to their classical hydrocarbon counterparts, so much so
that phosphorus has been coined “The Carbon Copy” in
the title of a recent book devoted to the subject.1a
Despite these similarities, low-coordinate phosphorus
ligands were largely thought of as chemical curiosities
that would have little practical application, a result of
the reactive nature of their P-C multiple bonds. This
view is quickly changing, as it is being realized that the
special electronic properties of cyclic and acyclic low-
coordinate phosphorus ligands can be harnessed in
transition metal complexes that show high activity and
selectivity in a variety of catalytic processes.1c
Of late, our research interests in low-coordination
phosphorus chemistry have stretched to the 1,3,5-
triphosphabenzene, P3C3But3, 1.6,7 Photoelectron spec-
troscopic studies have shown that the LUMOs of this
heterocycle in the complexes [M(CO)3(η6-P3C3But3)],
M ) Cr, Mo, or W, are lower in energy than in the cor-
Most relevant to the work reported here are com-
plexes of monophosphabenzenes (phosphinines). Ex-
amples incorporating kinetically stabilized phosphin-
ines, η1-P-coordinated to rhodium(I) fragments, have
been shown by Breit et al. to be highly active and
selective catalysts for the hydroformylation of olefins.2
(2) Breit, B.; Winde, R.; Mackewitz, T.; Paciello, R.; Harms, K. Chem.
Eur. J. 2001, 7, 3106.
(3) Knock, F.; Kremer, F.; Schmidt, U.; Zenneck, U.; Le Floch, P.;
Mathey, F. Organometallics 1996, 15, 2713.
(4) Le Floch, P.; Knoch, F.; Mathey, F.; Scholz, J.; Scholz, W.; Thiele,
K.-H.; Zenneck, U. Eur. J. Inorg. Chem. 1998, 119.
(5) Moores, A.; Mezailles, N.; Ricard, L.; Le Floch, P. Organometal-
lics 2005, 24, 508.
(6) Tabellion, F.; Nachbauer, A.; Leininger, S.; Peters, C.; Preuss,
F.; Regitz, M. Angew. Chem., Int. Ed. 1998, 37, 1233.
(7) Jones, C.; Waugh, M. Dalton Trans. 2004, 1971.
* To whom correspondence should be addressed. E-mail: jonesca6@
cardiff.ac.uk.
(1) For example: (a) Dillon, K. B.; Mathey, F.; Nixon, J. F. In
Phosphorus: The Carbon Copy; Wiley: Chichester, 1998. (b) Phospho-
rus-Carbon Heterocyclic Chemistry: The Rise of a New Domain;
Mathey, F., Ed.; Pergamon: Amsterdam, 2001. (c) Mathey, F. Angew.
Chem., Int. Ed. 2003, 42, 1578, and references therein.
10.1021/om0503489 CCC: $30.25 © 2005 American Chemical Society
Publication on Web 07/16/2005

Complexes of Group 8 and 9 Metals
Organometallics, Vol. 24, No. 17, 2005 4217
responding arene complexes, [M(CO)3(η6-C6H3But3)].8 As
a result, there is a greater degree of metal to ligand
back-bonding in the former, which leads to a stronger
coordination of this heterocycle. Despite this, only a
handful of spectroscopically characterized η6-P3C3But3
complexes (with group 6-8 metal fragments) are
known,8,9 and the only crystallographically character-
ized example has the ligand bridging two scandium(I)
centers, [{Sc(1,2,4-P3C2But3)}2(µ-η6-P3C3But3)].10 In ad-
dition to these complexes, several η1-complexes of 1 with
square planar Pt(II) fragments have been reported.11
One aim of this study was to extend the η6-coordination
chemistry of 1 to the preparation of group 8 and 9 metal
complexes, which could possibly lead to catalytic ap-
plications (cf. η6-phosphinine chemistry). It was also
thought of sufficient interest to prepare η5-triphos-
phacyclohexadienyl complexes of these metals for the
same purpose (cf. η5-phosphacyclohexadienyl chemis-
try). Such complexes are unknown, but the recently
reported syntheses of lithium triphosphacyclohexadienyl
salts, e.g., [Li][RP3C3But3], from 1 and alkyllithium
reagents, RLi, lent hope to this cause.12 It seemed these
salts could be reacted with transition metal halides to
give the desired complexes. The sometimes unexpected
results of our investigations in these areas are reported
herein.
Scheme 1
at 25 °C. The eliminated phosphinidene fragments in
the reaction gave rise to the known cyclophosphanes
(PMe)n,16 as evidenced by the presence of multiplet
signals centered at ca. δ 10 ppm in the 31P NMR
spectrum of the reaction mixture. It is noteworthy that
the lithium salt of 2 also eliminates PMe to give the
diphosphacyclopentadienyl anion, [1,3-P2C3But3]-, but
only on heating in THF at reflux for 3 h.17 Therefore, it
seems that coordination of 2 to iron significantly lowers
the energy barrier to this elimination process. We have
demonstrated similar chemistry with the p-block ele-
ment halides, TlCl, “GaI”, SnCl2, and PbCl2, which react
with 2 to give the corresponding mono- or bis(diphos-
phacyclopentadienyl) complexes [M(P2C3But3)n], n ) 1
or 2, in good yield.18 This is of synthetic importance
because, until recently,19 synthetic routes to the anion
[P2C3But3]- were difficult and low yielding, and its
homoleptic complexes are still rare.
In contrast to the reaction with FeCl2, when 2 was
reacted with [RuCl2(PPh3)3] in THF, the only product
recovered (ca. 60% yield) was the known triphosphacy-
clohexa-1,4-diene, 520 (Scheme 1). This is a recognized
hydrolysis product of 2, but cannot arise from the
presence of water in this case because the exclusion of
moisture from the reaction mixture was rigorous. In-
stead, it likely derives from hydrogen abstraction from
the THF solvent. Indeed, 5 was also produced in small
amounts in the iron reaction. It is not known why these
differences occur, but it is worth mentioning that we
have seen high yields of 5 arise from reactions of 2 with
group 4 to 7 metal halides.18
Results and Discussion
Group 8 Chemistry. Dimroth et al. reported the
synthesis of a range of bis(phosphacyclohexadienyl)iron
complexes by reaction of phosphacyclohexadienyl anions
with FeCl2 in 1985.13 A subsequent crystal structure
analysis of one of these sandwich complexes, [Fe(MePC5-
Ph2H2But)2], showed the heterocycles to be η5-coordi-
nated to the metal center.14 In an attempt to form a
hexaphospha analogue of these complexes, the triphos-
phacyclohexadienyl anion, [MeP3C3But3]-, 2,12 was re-
acted with FeCl2 in a 2:1 stoichiometry. The 31P{1H}
NMR spectrum of the reaction mixture suggested the
presence of several products. Upon workup, two were
isolated as crystalline solids in low to moderate yields,
viz., the known tetraphosphaferrocene, 3,15 and the
unusual organometallic cage complex, 4 (Scheme 1). It
seems likely that the intermediate in this reaction is
the expected complex [Fe(η5-MeP3C3But3)2], but this
subsequently eliminates 2 equiv of the phosphinidene
(PMe) to give 3 or undergoes an intramolecular phos-
phinidene transfer reaction to give 4. Both transforma-
tions must be facile, as following the reaction by 31P{1H}
NMR spectroscopy showed no evidence of the interme-
diate. In addition, both 3 and 4 had formed after 5 min
(8) Clendenning, S. B.; Green, J. C.; Nixon, J. F. J. Chem. Soc.,
Dalton Trans. 2000, 1507.
(9) Binger, P.; Stutzmann, S.; Stannek, J.; Gabor, B.; Mynott, R.
Eur. J. Inorg. Chem. 1999, 83.
The spectroscopic data for 4 are consistent with its
proposed structure. Most informative is its 31P{1H}
NMR spectrum, which displays four signals, one for the
(10) Arnold, P. L.; Cloke, F. G. N.; Hitchcock, P. B.; Nixon, J. F. J.
Am. Chem. Soc. 1996, 118, 7630.
(11) Clendenning, S. B.; Hitchcock, P. B.; Nixon, J. F. Chem.
Commun. 1999, 1377.
(12) Steinbach, J.; Renner, J.; Binger, P.; Regitz, M. Synthesis 2003,
1526.
(16) Berger, S.; Braun, S.; Kalinowski, H.-O. 31P NMR-Spectroskopie,
Vol. 3; Thieme: Stuttgart, 1991; p 140.
(13) Dave, T.; Berger, S.; Bilger, E.; Kaletsch, H.; Pebler, J.; Knecht,
J.; Dimroth, K. Organometallics 1985, 4, 1565.
(14) Baum, G.; Massa, W. Organometallics 1985, 4, 1574.
(15) Bartsch, R.; Cloke, F. G. N.; Green, J. C.; Matos, R. M.; Nixon,
J. F.; Suffolk, R. J.; Suter, J. L.; Wilson, D. J. J. Chem. Soc., Dalton
Trans. 2001, 1013.
(17) Steinbach, J.; Binger, P.; Regitz, M. Synthesis 2003, 2720.
(18) Brym, M. Ph.D. Thesis, Cardiff University, 2004.
(19) Cloke, F. G. N.; Hitchcock, P. B.; Nixon, J. F.; Wilson, D. J.
Organometallics 2000, 19, 219.
(20) Renner, J.; Bergstra¨sser, U.; Binger, P.; Regitz, M. Angew.
Chem., Int. Ed. 2003, 42, 1863.

4218 Organometallics, Vol. 24, No. 17, 2005
Francis et al.
Scheme 2
bond lengths are in the normal range,23 as are the Fe-P
bonds. Perhaps surprisingly, however, the dative P(1)-
Fe(1) bonds [2.1570(10) Å] are significantly shorter than
the phosphide P(2)-Fe(1) interaction [2.3013(13) Å].
To extend the η6-coordination chemistry of the tri-
phosphabenzene 1, and to generate potential precursors
for triphosphacyclohexadienyl complexes, 1 was reacted
with [Cp*Ru(NCMe)3]PF6 to give a good yield of the
desired complex, 6, after recrystallization from a dichlo-
romethane/hexane solution (Scheme 2). Surprisingly,
treatment of 1 with [Cp*Fe(NCMe)3]PF6 led to no
reaction, perhaps because the smaller covalent radius
of iron (1.16 Å) compared to that of ruthenium (1.24 Å)24
disfavors η6-coordination of the heterobenzene. When
6 was reacted with 1 equiv of MeLi, the first example
of a η5-triphosphacyclohexadienyl complex, 7, resulted
from nucleophilic attack at a P-center of the heterocycle.
This is favored over attack at the C-centers of the
heterocycle due to the polarization of its δ+P-Cδ- bonds.
The isolated yield from this reaction was low (ca. 10%),
and so an alternative synthetic route was devised
whereby [Cp*Ru(NCMe)3]PF6 was treated with the
anion 2 in a salt elimination reaction to give 7 in a 40%
isolated yield.
An attempt was also made to prepare the secondary
phosphine analogue of 7 by treating 6 with a 10-fold
excess of a hydride source, LiAlH4. However, after
aqueous workup, an isomer of the expected product
containing a saturated C-center within the heterocycle
was obtained, viz., 8. Although the isolated yield of this
compound was again low (22%), the 31P NMR spectrum
of the workup solution suggested 8 was the major
product. Presumably, this compound proved difficult to
crystallize in high yields because of its high solubility
in hexane. It is believed that 8 forms via its expected
secondary phosphine isomer, which undergoes a rapid
P to C hydrogen migration reaction. Closely related
migrations have been proposed for the conversion of the
1-H-phosphacyclohexadienyl complex [CpFe{η5-(H)PC5-
H2Ph3}] to its 2-H- and 3-H-isomers, [CpFe{η5-(H)(Ph)-
CPC4H2Ph2}] and [CpFe{η5-(H)2CPC4HPh3}], in the
presence of a hydride source.25 Complex 8 does not arise
Figure 1. Thermal ellipsoid plot (25% probability surface)
of the molecular structure of [(η5-1,3-P2C3But3)Fe{P2(PMe)2-
(CBut)3}] (4); hydrogen atoms are omitted for clarity. Fe-
(1)-P(1) 2.1570(10), Fe(1)-P(2) 2.3013(13), Fe(1)-P(4)
2.3287(10), Fe(1)-C(11) 2.254(5), Fe(1)-C(15) 2.217(3),
P(4)-C(11) 1.742(3), P(4)-C(15) 1.796(4), P(1)-C(3) 1.837-
(4), P(1)-C(2) 1.855(3), P(1)-C(1) 1.890(4), P(2)-C(2)
1.937(3), P(3)-C(1) 1.881(5), P(3)-C(2) 1.902(4), P(1)-Fe-
(1)-P(1)′ 74.53(5), P(1)-Fe(1)-P(2) 74.38(4), C(2)-P(1)-
C(1) 87.23(19), C(2)-P(1)-Fe(1) 100.99(11), C(1)-P(1)-
Fe(1) 98.54(11), C(2)-P(2)-C(2)′ 85.1(2), C(2)-P(2)-Fe(1)
93.69(11), C(1)-P(3)-C(2) 86.15(15), C(2)-P(3)-C(2)′ 87.0-
(2), P(3)-C(1)-P(1) 93.08(19), P(1)-C(1)-P(1)′ 87.4(2),
P(1)-C(2)-P(3) 93.54(16), P(1)-C(2)-P(2) 90.65(15), P(3)-
C(2)-P(2) 93.78(16). Symmetry transformation used to
generate equivalent atoms ′: x, -y+1/2, z.
heterocycle (δ 57.2 ppm) and three for the heterocubane
moiety. The signal for the two chemically equivalent
methylated P-centers, which are coordinated to the iron
atom, appears as a singlet, δ 49.1 ppm, while the other
tertiary P atom and the iron-coordinated phosphide
center resonate at δ 106.5 and 94.0 ppm, respectively.
These doublet signals (2JPP ) 18.0 Hz) are at relatively
low field due to the strain in the heterocubane fragment,
which gives rise to bonds of high p-character to these P
atoms. A similar, though more pronounced, phosphorus
deshielding has been reported for the tetraphosphacu-
bane, P4C4But4, δ 257 ppm.21
The molecular structure of this unusual complex is
depicted in Figure 1 (see also Table 1) and shows it to
contain a delocalized diphosphacyclopentadienyl ring η5-
coordinated to the iron center which forms one vertex
of a P4C3Fe heterocubane fragment. The geometry of
the heterocycle is similar to that in other complexes
incorporating it, e.g., [Cp*Fe(η5-P2C3But3)],22 while the
geometry of the heterocubane is strained, with intracage
angles ranging from 74.38(4) to 100.99(11)°. Its P-C
(21) Wettling, T.; Schneider, J.; Wagner, O.; Kreiter, C. G.; Regitz,
M. Angew. Chem., Int. Ed. Engl. 1989, 28, 1013.
(22) Muller, C.; Bartsch, R.; Fischer, A.; Jones, P. G.; Schmutzler,
R. J. Organomet. Chem. 1996, 512, 141.
(23) As determined by a survey of the Cambridge Crystallographic
Database, May 2005.
(24) Emsley, J. The Elements, 2nd ed.; Clarendon: Oxford, 1991.
(25) Nief, F.; Fischer, J. Organometallics 1986, 5, 877.

Complexes of Group 8 and 9 Metals
Organometallics, Vol. 24, No. 17, 2005 4219

4220 Organometallics, Vol. 24, No. 17, 2005
Francis et al.
Figure 3. Thermal ellipsoid plot (25% probability surface)
of the molecular structure of [Cp*Ru{η5-(MeP)P2C3But3}]
(7); hydrogen atoms are omitted for clarity. Ru(1)-C(1)
2.328(14), Ru(1)-P(1)′ 2.448(10), C(1)-P(1)′ 1.802(5), C(1)-
P(1)′′ 1.781(5), P(2)-C(1)′ 1.829(6), P(2)-C(1)′′ 1.841(6),
P(1)′-C(1)-P(1)′′ 116.4(3), C(1)-P(1)′-C(2)′′ 99.6(3), C(1)′-
P(2)-C(1)′′ 96.4(3). Symmetry transformations used to
generate equivalent atoms ′: y, z, x; ′′: z, x, y.
Figure 2. Thermal ellipsoid plot (25% probability surface)
of the cationic component of [Cp*Ru(η6-P3C3But3)][PF6] (6);
hydrogen atoms are omitted for clarity. Ru(1)-C(1) 2.341-
(4), Ru(1)-C(2) 2.353(4), Ru(1)-C(3) 2.357(4), Ru(1)-P(2)
2.4842(13), Ru(1)-P(1) 2.4857(15), Ru(1)-P(3) 2.4908(14),
P(1)-C(3) 1.748(5), P(1)-C(1) 1.750(5), P(2)-C(2) 1.744-
(4), P(2)-C(1) 1.756(5), P(3)-C(2) 1.753(5), P(3)-C(3)
1.756(5), C(3)-P(1)-C(1) 107.7(2), C(2)-P(2)-C(1) 107.0-
(2), C(2)-P(3)-C(3) 107.4(2), P(1)-C(1)-P(2) 132.6(3),
P(2)-C(2)-P(3) 132.9(3), P(1)-C(3)-P(3) 132.0(3).
these are longer than those in the free heterocycle [1.723
Å av],26 but shorter than the P-C bonds in the triphos-
phabenzene-bridged complex [{Sc(1,2,4-P3C2But3)}2(µ-
η6-P3C3But3)] [1.803 Å av].10 The distance from the Ru-
center to the heterocycle centroid [1.677 Å] is significantly
less than the Ru-Cp* ring centroid separation [1.864
Å].
The molecular structures of the triphosphacyclohexa-
dienyl complexes, 7 and 8, are shown in Figures 3 and
4, respectively. In both, the heterocycle and Cp* ligands
are η5-coordinated to the metal center, while the alkyl
substituents of the saturated heterocyclic centers are
in exo-positions. The P-C bonds within each delocalized
heteropentadienyl fragment are similar [7 1.781 Å av;
8 1.767 Å av] and comparable to those in 6. Moreover,
the Cp* centroid-Ru distances and the heteropentadi-
enyl fragment centroid-Ru separations are identical for
both complexes. viz., 1.886 and 1.674 Å, respectively.
Group 9 Chemistry. In attempts to prepare group
9-triphosphacyclohexadienyl complexes, a number of
reactions were carried out between the anion 2 and
group 9 halide complexes, e.g., CoCl2, [RhCl(COD)]2,
[IrCl(COD)]2, [RhCl(PPh3)3], and [Ir(CO)Cl(PPh3)2]. In
all cases, inseparable mixtures of products resulted,
which included the triphosphacyclohexadiene 5. It was
believed that an alternative route to such complexes
could involve the synthesis of η6-triphosphabenzene
complexes, which could then be treated with nucleo-
philes or other reagents, cf. the preparation of 7 and 8.
Le Floch et al. have recently demonstrated the effective-
ness of this approach with the synthesis of the first Rh-
from the aqueous workup of the reaction, as the 31P
NMR spectrum of the reaction mixture prior to quench-
ing shows 8 as the major product. In addition, treating
6 with 1 M HCl or NaOH solutions did not yield 8 but
gave compound mixtures of unknown identity.
The 31P{1H} NMR spectra for the complexes 6-8 are
compatible with their proposed formulations. That for
6 displays a singlet resonance for the heterocycle at δ
25.3 ppm, considerably upfield of the resonance for the
free triphosphabenzene, δ 257 ppm,6 but close to that
of related complexes, e.g., [Ru(η4-COD)(η6-P3C3But3)] δ
59.0 ppm.9 The spectra for 7 and 8 both exhibit AX2
systems, the signals for which in the spectrum of 7 [δ
-35.2 ppm (PA), 30.9 ppm (PX)] have shifted signifi-
cantly from those for the lithium salt of the free anion,
2 [δ -91.5 ppm (PA), 192.4 ppm (PX)].12 In addition, the
doublet signal in the spectrum of 8 [δ -111.4 ppm (PX)]
is at much higher field than the triplet resonance [δ 34.9
ppm (PA)]. Moreover, this signal is at higher field than
the region that might be expected for a coordinated P-C
multiple bond. There is a precedent for such a high-
field shift in the 31P NMR spectrum of the aforemen-
tioned 2-H-phospacyclohexadienyl complex, [CpFe{η5-
(H)(Ph)CPC4H2Ph2}], the phosphorus center of which
resonates at δ -173.0 or -150.1 ppm for its exo- and
endo-isomers, respectively.25
The structure of the cationic component of 6 is
depicted in Figure 2 and represents the first structural
characterization of a nonbridging η6-triphosphabenzene
complex. Its heterocycle is essentially planar, with P-C
bond lengths that are similar [1.751 Å av] and sugges-
tive of delocalization within the ring. Not surprisingly,
(26) Gleiter, R.; Lange, H.; Binger, P.; Stannek, J.; Kruger, C.;
Bruchmann, J.; Zenneck, U.; Kummer, S. Eur. J. Inorg. Chem. 1998,
1619.

Complexes of Group 8 and 9 Metals
Organometallics, Vol. 24, No. 17, 2005 4221
Scheme 3
and an unidentified mixture of other products. Further-
more, treating a mixture of 1 and [RhCl(PPh3)2]2 with
Na[BArf4] led to no Rh coordination of the triphospha-
benzene, most probably because of the greater steric
demands of two PPh3 ligands relative to one COD
ligand.
We have begun to examine the further chemistry of
9 and have made some interesting observations. First,
the triphosphabenzene ligand is not displaced and the
complex remains intact when it is treated with large
excesses of arenes, e.g., benzene, or two- or four-electron
donors such as PPh3, CO, or bipyridine. This is in
contrast to corresponding [Rh(COD)(arene)]+ complexes,
the arene ligands of which are generally readily dis-
placed.27 In the case of 9, two factors probably hinder
displacement of the triphosphabenzene: the steric
protection this bulky ligand affords the rhodium center
and a stronger interaction with the rhodium center than
is normal for arenes. The latter is reasonable consider-
ing the known lower energy LUMOs (and greater
Mfring back-bonding) of 1 relative to those of arenes
in their respective transition metal complexes.8
Figure 4. Thermal ellipsoid plot (25% probability surface)
of the molecular structure of [Cp*Ru{η5-P3C2But2(CHBut)}]
(8); methyl hydrogen atoms are omitted for clarity. Ru(1)-
C(2) 2.264(3), Ru(1)-C(1) 2.274(3), Ru(1)-P(3) 2.4066(9),
Ru(1)-P(1) 2.4118(9), Ru(1)-P(2) 2.4599(10), P(1)-C(1)
1.778(3), P(1)-C(3) 1.857(3), P(2)-C(1) 1.757(3), P(2)-C(2)
1.765(3), P(3)-C(2) 1.768(4), P(3)-C(3) 1.850(3), C(1)-P(1)-
C(3) 107.51(14), C(1)-P(2)-C(2) 104.59(14), C(2)-P(3)-
C(3) 108.18(13), P(2)-C(1)-P(1) 127.47(17), P(2)-C(2)-
P(3) 127.23(16), P(3)-C(3)-P(1) 96.63(15).
The coordination of 1 also seems to activate it toward
attack at its P-centers, as has been found to be the case
with monophosphabenzenes. An excellent example of
this is the fact that in the uncoordinated state 1 is
resistant toward reaction with water or alcohols.11
However, one P-center of complex 9 readily reacts at 0
°C with an excess of water or ethanol to give the
triphosphacyclohexadienyl complexes 10 and 11 as the
only phosphorus-containing products in the reaction
mixtures (Scheme 3). These 1,1-addition reactions are
completely analogous to those involving the aforemen-
tioned monophosphabenzene-rhodium(I) cationic com-
plexes reported by Le Floch et al.27 A number of other
reactions were carried out between 9 and precursors
that could potentially oxidatively add to its rhodium
center or attack a phosphorus center of its heterocyclic
ligand. Little success has so far been had here, as 9 was
found to be unreactive to excesses of H2 (5 atm/50 °C)
or MeI, while reactions with excess anhydrous HCl,
triflic acid, or Et3SiH led to intractable mixtures of
phosphorus-containing products. In addition, and in
contrast to other cationic Rh(I) complexes,30 complex 9
was found to be unreactive toward the oxidative addi-
tion of bis(catecholato)diborane.
(I) and Ir(I) η6-monophosphabenzene complexes, e.g.,
[Rh(η4-COD){η6-PC5H(SiMe3)2(Ph)2}]+, which undergo
1,1-addition reactions with alcohols or water to give the
first examples of Rh(I) and Ir(I) phosphacyclohexadienyl
complexes, e.g., [Rh(η4-COD){η5-(H)(EtO)PC5H(SiMe3)2-
(Ph)2}]+.27 It seemed reasonable that similar chemistry
could be carried out starting with the triphosphaben-
zene 1.
Initial attempts to prepare the cationic complex [Rh-
(COD)(P3C3But3)]+ by treating a mixture of [RhCl-
(COD)]2 and P3C3But3 in CH2Cl2 with NaBPh4 met with
failure and led only to the known complex [(COD)Rh-
{η6-(C6H5)BPh3}]28 and unreacted triphosphabenzene.
This presumably occurs as one phenyl arm of the
[BPh4]- anion successfully competes with the triphos-
phabenzene for Rh coordination. As a result, the sodium
salt of the more weakly coordinating anion, [BArf4]-,
Arf ) C6H3(CF3)2-3,5, was used in this reaction. Al-
though this anion can also coordinate to the [Rh(COD)]+
fragment,29 in this case it does not and the first
triphosphabenzene-group 9 complex, 9, was formed
cleanly and in good yield (Scheme 3). Surprisingly,
attempts to prepare the iridium analogue of this com-
plex by an analogous procedure led only to approxi-
mately 50% of the triphosphabenzene being consumed
The 31P{1H} NMR spectrum of the triphosphabenzene
complex 9 displays a singlet resonance, δ 108.3 ppm,
which is considerably downfield of that for the ruthe-
nium complex, 6 (δ 25.3 ppm), but close to the position
of the signals in related monophosphabenzene-rho-
(27) Doux, M.; Ricard, L.; Mathey, F.; Le Floch, P.; Mezailles, N.
Eur. J. Inorg. Chem. 2003, 687.
(28) Schrock, R. R.; Osborn, J. A. Inorg. Chem. 1970, 9, 2339.
(29) Powell, J.; Lough, A.; Saeed, T. J. Chem. Soc., Dalton Trans.
1997, 4137.
(30) Morgan, J. B.; Miller, S. P.; Morken, J. P. J. Am. Chem. Soc.
2003, 125, 8702.

4222 Organometallics, Vol. 24, No. 17, 2005
Francis et al.
Figure 5. Thermal ellipsoid plot (25% probability surface) of the structure of the cationic component of [(COD)Rh(η6-
P3C3But3)][BArf4] (9); hydrogen atoms are omitted for clarity. Rh(1)-C(1) 2.321(4), Rh(1)-C(2) 2.372(4), Rh(1)-C(3) 2.445-
(4), Rh(1)-P(2) 2.5220(11), Rh(1)-P(3) 2.5293(12), Rh(1)-P(1) 2.5683(12), P(1)-C(3) 1.739(4), P(1)-C(1) 1.766(4), P(2)-
C(3) 1.758(4), P(2)-C(2) 1.761(4), P(3)-C(2) 1.745(4), P(3)-C(1) 1.753(4), C(3)-P(1)-C(1) 107.13(18), C(3)-P(2)-C(2)
109.48(18), C(2)-P(3)-C(1) 106.51(18), P(3)-C(1)-P(1) 134.1(2), P(3)-C(2)-P(2) 131.6(2), P(1)-C(3)-P(2) 131.0(2).
dium(I) complexes, e.g., δ 101.4 ppm for [Rh(COD)-
{η6-PC5Me2(SiMe3)2}]+.27 However, unlike the latter
complexes, which generally exhibit small 1JRhP couplings
(ca. 10 Hz), no such coupling was observed in the
spectrum of 9. Similarly no Rh-P couplings were seen
in the 31P NMR spectra of the 1,1-addition products 10
and 11, both of which exhibit AX2 spin systems with
the doublet signals (δ 117.0 and 121.1 ppm, respectively)
close to the position of the resonance in the spectrum
of 9. The triplet resonances (δ -13.0 and -8.5 ppm,
respectively) are at considerably higher field and exhibit
large 1JPH couplings (560 and 578 Hz, respectively). The
presence of P-H bonds in these complexes was con-
firmed by the observation of characteristic stretching
absorptions in their IR spectra.
An examination of the structure of the cationic
component of 9 (Figure 5) showed it to have no signifi-
cant interaction with the [BArf4]- anion and to contain
an effectively planar triphosphabenzene ligand η6-
coordinated to the rhodium center. The intracyclic P-C
bond lengths are similar [1.754 Å av] and close to those
in the ruthenium complex, 6. There is, however, a
significant difference between the M-heterocycle cen-
troid distances in the two compounds [6 1.677 Å; 9 1.734
Å], which mirrors the expected smaller covalent radius
of Ru(II) relative to Rh(I).
η5-coordinated to the rhodium centers, as has been seen
in the structure of the related iridium(I) complex
[Ir(COD){η5-(H)(EtO)PC5(Me)2(SiMe3)2}]+.27
Conclusions
In summary, reactions of the 1,3,5-triphosphabenzene
P3C3But3 or the triphosphacyclohexadienyl anion [MeP3-
C3But3]- toward a series of group 8 and 9 halide
complexes have been investigated. These have given rise
to the first examples of triphosphacyclohexadienyl-
transition metal complexes and the first structurally
characterized complexes containing a nonbridging η6-
coordinated triphosphabenzene ligand. The facility of
heterocycle transformations at metal centers has been
demonstrated with the nucleophilic attack of a triphos-
phabenzene within the coordination sphere of a ruthe-
nium(II) fragment to yield either 1-R- or 2-H-triphos-
phacyclohexadienyl complexes. Formal 1,1-additions of
water or ethanol to a phosphorus center of the same
heterobenzene, but within the coordination sphere of a
rhodium(I) fragment, have yielded related complexes
containing zwitterionic triphosphacyclohexadienyl lig-
ands. Evidence has also been presented to show that in
the coordination sphere of an iron(II) center the triphos-
phacyclohexadienyl ligand [MeP3C3But3]- readily elimi-
nates a phosphinidene fragment, PMe, to give η5-1,3-
diphosphacyclopentadienyl complexes. Taken as a whole,
this study has highlighted both similarities and differ-
ences in the reactivity of triphosphaheterocycle com-
plexes, compared to that of their monophospha coun-
terparts. Considering the emerging importance of the
latter in a range of catalytic applications, there is much
scope to extend investigations into the chemistry of the
The structures of the cations of the closely related
complexes 10 and 11 are very similar and are depicted
in Figures 6 and 7, respectively. Both display significant
positional disorder within the heterocycles, which,
although successfully modeled, does not allow definitive
comment to be made on the geometries of those hetero-
cycles. Saying that, it is clear that the diphosphapen-
tadienyl fragments of each are essentially planar and

Complexes of Group 8 and 9 Metals
Organometallics, Vol. 24, No. 17, 2005 4223
Figure 6. Thermal ellipsoid plot (25% probability surface)
of the structure of the cationic component of [(COD)Rh-
{η5-[(H)(HO)P]P2C3But3}][BArf4] (10); methyl hydrogen
atoms are omitted for clarity. Rh(1)-C(3) 2.300(4), Rh(1)-
C(2) 2.331(4), Rh(1)-C(1) 2.410(4), Rh(1)-P(3) 2.4389(13),
Rh(1)-P(2) 2.4918(18), P(1)-O(1) 1.565(5), P(1)-C(3) 1.727-
(4), P(1)-C(1) 1.749(4), P(2)-C(1) 1.731(4), P(2)-C(2)
1.739(4), P(3)-C(2) 1.766(4), P(3)-C(3) 1.772(4), O(1)-
P(1)-C(3) 115.6(2), O(1)-P(1)-C(1) 115.0(2), C(3)-P(1)-
C(1) 108.5(2), C(1)-P(2)-C(2) 109.4(2), C(2)-P(3)-C(3)
104.67(19), P(2)-C(1)-P(1) 120.6(2), P(2)-C(2)-P(3) 132.9-
(2), P(1)-C(3)-P(3) 124.2(3).
Figure 7. Thermal ellipsoid plot (25% probability surface)
of the structure of the cationic component of [(COD)Rh-
{η5-[(H)(EtO)P]P2C3But3}][BArf4] (11); carbon bound hy-
drogen atoms are omitted for clarity. Rh(1)-C(1) 2.312(4),
Rh(1)-C(3) 2.361(4), Rh(1)-C(2) 2.389(4), Rh(1)-P(3)
2.4898(14), Rh(1)-P(2) 2.520(1), P(1)-O(1) 1.579(5), P(1)-
C(1) 1.628(5), P(1)-C(3) 1.801(5), P(2)-C(2) 1.781(5), P(2)-
C(1) 1.857(6), P(3)-C(2) 1.752(5), P(3)-C(3) 1.753(4), O(1)-
P(1)-C(1) 119.0(3), O(1)-P(1)-C(3) 116.2(2), C(1)-P(1)-
C(3) 108.8(2), C(2)-P(2)-C(1) 99.9(2), C(2)-P(3)-C(3)
108.0(2), P(1)-C(1)-P(2) 125.1(3), P(3)-C(2)-P(2) 135.8-
(3), P(3)-C(3)-P(1) 117.4(3).
former. Such studies are ongoing in our laboratory and
will be reported in due course.
Preparation of [(η5-1,3-P2C3But3)Fe{P2(PMe)2(CBut)3}]
(4). P3C3But3 (0.150 g, 0.5 mmol) was dissolved in a mixture
of THF (10 mL)/tmeda (1 mL) and cooled to -78 °C. To this
solution was added a diethyl ether solution of MeLi (0.55
mmol) over 5 min. After 30 min, the reaction mixture was
warmed to ambient temperature and stirred for 1 h. Volatiles
were removed to give an orange powder. This was dissolved
in THF (5 mL) and added over 5 min to a suspension of FeCl2
(32 mg, 0.25 mmol) in THF (15 mL) at -78 °C. The resultant
solution was warmed to room temperature and stirred over-
night. Volatiles were then removed in vacuo, and the residue
was extracted with hexane and filtered. Concentration and
slow cooling of the filtrate to -30 °C afforded 4 as deep red
Experimental Section
General Methods. All manipulations were carried out
using standard Schlenk and glovebox techniques under an
atmosphere of high-purity argon. Diethyl ether, hexane, and
THF were distilled over Na/K alloy, while CH2Cl2 was distilled
over CaH2. 1H, 13C, and 31P NMR spectra were recorded on
either Bruker DPX400 or JEOL Eclipse 300 spectrometers in
deuterated solvents and were referenced to the residual 1H or
13C resonances of the solvent used (1H and 13C NMR) or
external 85% H3PO4, δ 0.0 ppm (31P NMR). Mass spectra were
recorded using a VG Fisons Platform II instrument operating
under APCI conditions or were obtained from the EPSRC
National Mass Spectrometric Service at Swansea University.
IR spectra were recorded using a Nicolet 510 FT-IR spectrom-
eter as Nujol mulls between NaCl plates. Melting points were
determined in sealed glass capillaries under argon and are
uncorrected. Microanalyses were obtained from Medac Ltd.
Reproducible microanalyses for compounds 4 and 6-8 could
not be obtained, but the NMR spectra of bulk samples of these
compounds suggested a purity of greater than 95% in each
case (see Supporting Information). P3C3But3,6 [Cp*Ru(NCMe)3]-
[PF6],31 [Rh(COD)Cl]2,32 and Na[BArf4]33 were synthesized by
literature procedures. All other reagents were used as received.
1
crystals (46 mg, 27%). Mp: 286-288 °C (dec). H NMR (400
MHz, C6D6, 300 K): δ 0.93 (s, 9 H, But), 1.07 (s, 18 H, 2 ×
But), 1.25 (s, 9 H, But), 1.75 (s, 18 H, 2 × But), 2.42 (m, 6 H,
PMe). 31P{1H} NMR (121.7 MHz, C6D6, 300 K): δ 49.1 (s, 2 ×
PMe), 57.2 (s, P2C3But3), 94.0 (d, P(CBut)2Fe, 2JPP ) 18.0 Hz),
106.5 (br d, P(CBut)3, 2JPP ) 18.0 Hz). IR ν/cm-1 (Nujol): 1262-
(s), 1202(s), 919(sh), 883(sh), 829(s), 775(w), 760(w), 721(s),
669(w). MS/EI: m/z 686 [M+, 100%]. Acc. Mass EI+: calc for
C32H60FeP6 686.2465, found 686.2456. NB: after crystallization
of 4 from the reaction mixture, chromatography of the mother
liquor (silica gel/hexane) yielded [Fe(η5-P2C3But3)2], 3, as a
blue-green solid in 17% yield. The NMR data for this sample
were found to be in agreement with those previously re-
ported.15
Preparation of [Cp*Ru(η6-P3C3But3)][PF6] (6). To a
solution of [Cp*Ru(NCMe)3][PF6] (0.30 g, 0.59 mmol) in
dichloromethane (15 mL) was added a solution of P3C3But3
(0.185 g, 0.62 mmol) in dichloromethane (5 mL) over 5 min at
(31) Schrenk, J. L.; McNair, A. M.; McCormick, F. B.; Mann, K. R.
Inorg. Chem. 1986, 25, 3501.
(32) Giordano, G.; Crabtree, H. Inorg. Synth. 1990, 28, 88.
(33) Reger, D.; Wright, T.; Little, C.; Lambda, J.; Smith, M. Inorg.
Chem. 2001, 40, 3810.

4224 Organometallics, Vol. 24, No. 17, 2005
Francis et al.
0 °C. The resultant solution became orange and was stirred
for 4 h at ambient temperature. Volatiles were removed in
vacuo, and the residue was washed with hexane and dried in
vacuo, leaving a pale yellow solid. This was dissolved in CH2-
Cl2 (1 mL), and the solution slowly diffused into hexane to give
6 as yellow-orange crystals (220 mg, 55%). Mp: 220-222 °C
(dec). 1H NMR (400 MHz, CD2Cl2, 300 K): δ 1.60 (s, 27 H,
But), 1.89 (s, 15 H, Me). 13C NMR (101.6 MHz, CD2Cl2, 300
K): δ 12.7 (Me), 35.4 (C(CH3)3, 3JPC ) 12.7 Hz), 43.5 (C(CH3)3,
2JPC ) 23.0 Hz), 101.7 (C(CH3)), CP resonance not observed.
31P{1H} NMR (121.7 MHz, CD2Cl2, 300 K): δ 25.3 (s, P3C3-
8H, o-ArH). 13C NMR (101.6 MHz, CD2Cl2, 300 K): δ 31.6 (br
s, CH2), 35.1 (t,3JPC ) 13.1 Hz, CH3), 44.8 (m, C(CH3)3), 84.1
(d, CH, 1JRhC ) 9.0 Hz), 117.5 (s, p-ArC), 124.6 (q, CF3, 1JCF
)
2
272.2 Hz), 128.7 (q, m-ArC, JCF ) 31.5 Hz), 134.8 (s, o-ArC),
1
161.8 (q, ipso-ArC, JBC ) 50.1 Hz), 163.5 (m, CP). 31P{1H}
NMR (121.7 MHz, CD2Cl2, 300 K): δ 108.3 (s, P3C3But3). 11B-
{1H}NMR (94.4 MHz, CD2Cl2, 300 K): δ -7.6. 19F{1H} NMR
(282.8 MHz, CD2Cl2, 300 K): δ -62.71. IR ν/cm-1 (Nujol):
1353(s), 1279(s), 1161(s), 888(s), 837(m). MS/EI: m/z 511
[M - BArf4, 13%]; 300 [P3C3But3+, 100%]. Anal. Calc for C55-
H51BF24P3Rh: C 48.06, H 3.74. Found: C 47.23, H 3.70.
Preparation of [(COD)Rh{η5-[(H)(HO)P]P2C3But3}]-
[BArf4] (10). To a solution of 9 (150 mg, 0.11 mmol) in
dichloromethane (10 mL) at -78 °C was added H2O (10 µL,
0.55 mmol). After 10 min the solution was warmed to room
temperature and stirred overnight. Volatiles were then re-
moved in vacuo, and the residue was washed with hexane,
extracted into dichloromethane (1 mL), and layered with
hexane (10 mL) to give red-brown crystals of 10 after 2 days
1
But3), -143.9 (sept., PF6, JPF ) 712 Hz). IR ν/cm-1 (Nujol):
1078(s), 1021(s), 838(s), 721(m). MS/APCI: m/z 555 [Cp*Ru-
(P3C3But3F)+, 25%], 536 [M - PF6, 40%]. Acc. Mass ES+: calc
for M - PF6 (C25H42P3Ru) 537.1537, found 537.1543.
Preparation of [Cp*Ru{η5-(MeP)P2C3But3}] (7). P3C3-
But3 (0.150 g, 0.5 mmol) was dissolved in a mixture of THF
(10 mL)/tmeda (1 mL) and cooled to -78 °C. To this solution
was added a diethyl ether solution of MeLi (0.55 mmol) over
5 min. After 30 min, the reaction mixture was warmed to
ambient temperature and stirred for 1 h. Volatiles were
removed in vacuo to give an orange powder. This was dissolved
in THF (10 mL) and the solution added to a suspension of
[Cp*Ru(CH3CN)3][PF6] (0.252 g, 0.5 mmol) in THF (15 mL) at
-78 °C. The mixture was warmed to ambient temperature and
stirred overnight. Volatiles were then removed in vacuo, and
the residue was extracted with hexane (5 mL). Placement of
this extract at -30 °C overnight yielded 7 as yellow crystals
(110 mg, 40%). Mp: 118-121 °C (dec). 1H NMR (400 MHz,
1
(0.09 g, 52%). Mp: 86-92 °C (dec). H NMR (400 MHz, CD2-
Cl2, 300 K): δ 1.33 (s, 9 H, But), 1.45 (s, 18 H, But), 2.21 (m,
1
8 H, CH2), 4.62 (s, 4H, CH), 7.42 (d, 1 H, PH, JPH ) 560 Hz)
7.45 (s, 4 H, p-ArH), 7.62 (s, 8H, o-ArH), OH resonance not
observed. 13C NMR (101.6 MHz, CD2Cl2, 300 K): δ 31.5 (br s,
CH2), 33.0 (m, PACC(CH3)3PX), 34.4 (t,3JPC ) 11.1 Hz, PXCC-
(CH3)3PX), 39.4 (m, PACC(CH3)3PX), 43.2 (m, PXCC(CH3)3PX),
1
80.3 (d, CH, JRhC ) 12.1 Hz), 117.5 (s, p-ArC), 124.6 (q, CF3,
2
1JCF ) 272.2 Hz), 128.9 (q, m-ArC, JCF ) 31.1 Hz), 134.8 (s,
o-ArC), 161.5 (q, ipso-ArC, 1JBC)50.4 Hz), PC resonances not
observed. 31P{1H} NMR (121.7 MHz, CD2Cl2, 300 K): δ 117.3
2
C6D6, 300 K): δ 0.62 (d, 3 H, PMe, JPH ) 6.1 Hz), 1.48 (s, 18
H, But), 1.60 (s, 15 H, Me), 1.64 (s, 9 H, But). 13C NMR (101.6
(d, Px, JPP ) 14.9 Hz), -13.1 (t, PA, JPP ) 14.9 Hz). 11B{1H}
NMR (94.4 MHz, CD2Cl2, 300 K): δ -7.6. 19F{1H} NMR (282.8
MHz, CD2Cl2, 300 K): δ -62.73. IR ν/cm-1 (Nujol): 2358(m),
P-H. MS/EI: m/z 529 [M - BArf4, 100%]. Acc. Mass ES+: calc
for M - BArf4 (C23H41OP3Rh) 529.1420, found 529.1424. Anal.
Calc for C55H53BF24OP3Rh (vacuum-dried sample): C 47.44,
H 3.84. Found: C 47.02, H 3.78.
2
2
MHz, C6D6, 300 K): δ 10.1 (PMe), 12.7 (Me), 31.9 (m, PACC-
3
(CH3)3PX), 35.5 (tr, PXCC(CH3)3PX, JPC ) 13.3 Hz), 38.9 (m,
PACC(CH3)3PX), 41.5 (tr, PXCC(CH3)3PX, 2JPC ) 19.6 Hz), 93.0
(C(CH3)), CP resonances not observed. 31P{1H} NMR (121.7
2
MHz, C6D6, 300 K): δ 30.9 (d, PX, JPP ) 8.9 Hz), -35.2 (tr,
2
PA, JPP ) 8.9 Hz). IR ν/cm-1 (Nujol): 1261(s), 1213(s), 1022-
(s), 844(m). MS/APCI: m/z 553 [M+, 100%]. Acc. Mass ES+:
calc for M+ (C26H45P3Ru) 553.1850, found 553.1851.
Preparation of [(COD)Rh{η5-[(H)(EtO)P]P2C3But3}]-
[BArf4] (11). To a solution of 9 (150 mg, 0.11 mmol) in
dichloromethane (10 mL) at -78 °C was added ethanol (30
µL, 0.55 mmol). After 10 min the solution was allowed to warm
to room temperature and stirred overnight. Volatiles were then
removed in vacuo, and the residue was washed with hexane,
extracted into dichloromethane (1 mL), and layered with
hexane (10 mL) to give red-brown crystals of 11 after 2 days
(0.10 g, 58%). Mp: 166-168 °C (dec). 1H NMR (400 MHz, CD2-
Cl2, 300 K): δ 1.04 (tr, 3 H, CH3, 3JHH ) 7.1 Hz), 1.34 (s, 9 H,
But), 1.43 (s, 18 H, But), 2.23 (m, 8 H, CH2), 3.05 (q, 2 H, OCH2,
3JHH ) 7.1 Hz), 4.67 (s, 4 H, CH), 7.45 (s, 4 H, p-ArH), 7.60 (d,
1 H, PH, 1JPH ) 578 Hz), 7.63 (s, 8 H, o-ArH). 13C NMR (101.6
MHz, CD2Cl2, 300 K): δ 15.1 (br, CH3), 31.5 (br s, CH2), 32.7
(m, PACC(CH3)3PX), 34.6 (t,3JPC ) 11.1 Hz, PXCC(CH3)3PX), 39.4
(m, PACC(CH3)3PX), 43.4 (m, PXCC(CH3)3PX), 62.7 (br, OCH2),
Preparation of [Cp*Ru{η5-P3C2But2(CHBut)}] (8). To a
suspension of 6 (0.22 g, 0.33 mmol) in diethyl ether (10 mL)
was added a solution of LiAlH4 (0.l2 g, 3 mmol) in diethyl ether
(25 mL). The resultant mixture was stirred overnight, after
which it was quenched with an excess of degassed 1 M HCl
(aq). A solution of degassed NaHCO3 (aq) was then added with
rapid stirring until the pH was approximately 8. The ether
layer was separated, dried over MgSO4, and evaporated. The
residue was extracted with hexane (10 mL) and filtered, and
the filtrate was placed at -30 °C overnight to give 8 as yellow
1
crystals (39 mg, 22%). Mp: 147-149 °C. H NMR (400 MHz,
C6D6, 300 K): δ 0.84 (s, 9 H, But), 1.43 (s, 18 H, But), 1.77 (s,
15 H, Me), methine resonance obscured. 13C NMR (101.6 MHz,
C6D6, 300 K): δ 12.1 (Me), 28.5 (br s, PXCC(CH3)3PX), 31.8 (tr,
PXCC(CH3)3PX, 1JPC ) 38.9 Hz), 34.1 (m, PACC(CH3)3PX), 35.6
(br s, PXCC(CH3)3PX), 39.9 (m, PACC(CH3)3PX), 95.6 (C(CH3)),
PACPX resonance not observed. 31P NMR (121.7 MHz, C6D6,
1
80.9 (d, CH, JRhC ) 9.2 Hz), 117.5 (s, p-ArC), 124.6 (q, CF3,
2
1JCF ) 272.5 Hz), 129.0 (q, m-ArC, JCF ) 31.4 Hz), 134.8 (s,
o-ArC), 161.0 (q, ipso-ArC, 1JBC ) 50.4 Hz), PC resonances not
observed. 31P{1H} NMR (121.7 MHz, CD2Cl2, 300 K): δ 121.7
2
300 K): δ 34.9 (tr, PA, JPP ) 24.0 Hz), -111.4 (d of d, PX,
2
2
2
2JPP ) 24.0 Hz, JPH ) 15.0 Hz). IR ν/cm-1 (Nujol): 1260(s),
(d, Px, JPP ) 18.2 Hz), -8.4 (t, PA, JPP ) 18.2 Hz). 11B{1H}
NMR (94.4 MHz, CD2Cl2, 300 K): δ -7.6. 19F{1H} NMR (282.8
MHz, CD2Cl2, 300 K): δ -62.75. IR ν/cm-1 (Nujol): 2363(m),
P-H. MS/EI: m/z 557 [M - BArf4, 86%]. Acc. Mass EI+: calc
for M - BArf4 (C25H44OP3Rh) 557.1733, found 557.1730. Anal.
calc for C57H57BF24OP3Rh (vacuum-dried sample): C 48.19, H
4.04. Found: C 47.83, H 3.81.
1095(s), 1023(s), 802(m). MS/APCI: m/z 538 [M+, 100%]. Acc.
Mass EI+: calc for M+ (C25H43P3Ru) 538.1616, found 538.1609.
Preparation of [(COD)Rh(η6-P3C3But3)][BArf4] (9). To
a mixture of P3C3But3 (0.15 g, 0.50 mmol) and Na[BArf4] (0.44
g, 0.50 mmol) in dichloromethane (5 mL) at 25 °C was added
a solution of [Rh(COD)Cl]2 (0.12 g, 0.25 mmol) in dichlo-
romethane (10 mL). The resultant suspension was stirred
overnight, whereupon volatiles were removed in vacuo. The
residue was washed with hexane, extracted into dichlo-
romethane (1 mL), and layered with hexane to give red-brown
crystals of 9 over 2 days (0.31 g, 46%). Mp: 122-126 °C (dec).
1H NMR (400 MHz, CD2Cl2, 300 K): δ 1.54 (s, 27 H, But), 2.16
(m, 8 H, CH2), 4.84 (s, 4 H, CH), 7.45 (s, 4 H, p-ArH), 7.61 (s,
X-ray Crystallography. Crystals of 4 and 6-11 suitable
for X-ray structural determination were mounted in silicone
oil. Crystallographic measurements were made using a Nonius
Kappa CCD diffractometer using a graphite monochromator
with Mo KR radiation (λ ) 0.71073 Å). The data were collected
at 150 K, and the structures were solved by direct methods
and refined on F2 by full matrix least squares (SHELX97)34

Complexes of Group 8 and 9 Metals
Organometallics, Vol. 24, No. 17, 2005 4225
using all unique data. All non-hydrogen atoms are anisotropic
with hydrogen atoms included in calculated positions (riding
model). Crystal data, details of data collections, and refinement
are given in Table 1.
Union (visiting studentship for C.H.) for financial sup-
port. Mr. Arnim Eyssler is also thanked for reproducing
the syntheses of 3 and 4. We also gratefully acknowledge
the EPSRC Mass Spectrometry Service, Swansea.
Supporting Information Available: Crystallographic
data as CIF files for 4 and 6-11; 1H and 31P{1H} NMR spectra
for compounds 4 and 6-8. This material is available free of
charge via the Internet at http://pubs.acs.org.
Acknowledgment. We thank the EPSRC (partial
studentship for R.P.R. and postdoctoral fellowship for
M.D.F.) and the ERASMUS scheme of the European
(34) Sheldrick, G. M. SHELX-97; University of Go¨ttingen, 1997.
OM0503489